Lens Focal Length and Perspective

[dancqu]

Correct me if I'm wrong. A normal lens gives
a perspective near that of the eye. Dan

 

[Ian Grant]

All lenses give exactly the same perspective, if they are in the same position and shooting the same subject, regardless of focal length. Only the angle of view differs, and the magnification of the image on the film.

Perspective is purely a function of the distance from the object, so go in close with a WA and you get a very different perspective from moving a way with a telephoto.

Ian

 

[panchro-press]

To add to Mr. Grant's reply, a normal lens give an angle of view which approximates the view of the human eye.

Dave

 

[Justin Cormack]

Its not clear what the view of the human eye is - the eyes move, we have peripheral vision (that is different from central vision), we have a more pano view in a way, but then we move our heads and make something different. And we dont have sharp edges. A photo doesnt look like what we see.

 

[dancqu]

To add to Mr. Grant's reply, a normal lens give an angle of
view which approximates the view of the human eye. Dave

Perhaps perspective is not the correct word. I've in mind
two classic examples. One, the vanishing point associated
with railroad tracks, and two, the compression or lack of
when imaging telephone polls.

With a normal lens I expect near and far size relationships
to be about as I see them. Dan

 

[Ray Heath]

c'mon guys, why make it complicated

yes dan, in practice a 'normal' lens gives about the view of human eyes, not peripheral, just the central, about 50 degrees

be aware though that each camera format (size) has it's own 'normal' focal length which can be roughly deduced by measuring the diagonal of the format, e.g 50mm for a 35mm film camera, 300mm for an 8x10 inch camera

the focal length is an optical measure that doesn't change, whether it be normal, wide or telephoto depends on how much of the projected image is recorded by the format being used

Ray

 

[RobC]

Perhaps perspective is not the correct word. I've in mind
two classic examples. One, the vanishing point associated
with railroad tracks, and two, the compression or lack of
when imaging telephone polls.

With a normal lens I expect near and far size relationships
to be about as I see them. Dan

If you want the same as your eye, then find out what the focal length of the lense in your eye is and use the same on your camera.

 

[RobC]

no lense works like your eye.
Take two matches and place them an inch apart about two meters from you.
Then focus your eyes on one of them. Then in your peripheral vision see if the other match which is only an inch away is in focus. It won't be. Now what does that tell you about the difference between your eye, a lense, and what angle of view really means in relation to how we see.

 

[Larry.Manuel]

on-line photography course

There's a long, detailed, paid module on rogerandfrances.com called:

"perspective and the illusion of depth"

My money was very well spent on the subscription there. I learned about it from APUG.

Good luck.

L.

 

[RobC]

I just looked up the focal length of the eye and it is infact between approx 16mm and 17mm. It adjusts focal length by changing the shape of the lense as it can't move the lenses further away from the retina to focus.

So if you want the same near far relationship as the eye, then you need a 16 or 17mm lense.
Look straight ahead and hold your arms to the side. You can infact see them in your peripheral vision giving you almost 180deg angle of view but not in focus. So a 16 or 17mm lense should be quite close to that.

BUT don't forget that the brain processes the image and that can make things look very different from reality. For example, the image is upside down on the retina but you don't see it that way.

p.s. I realise that this appears to be going off at a tangent but it should serve to illustrate that there is no such thing as a camera lense which will give you the same as the eye sees.

 

[Ray Heath]

... approx 16mm and 17mm ...

interesting rob

but what about sensor size and angle of view

 

[DannL]

c'mon guys, why make it complicated

yes dan, in practice a 'normal' lens gives about the view of human eyes, not peripheral, just the central, about 50 degrees

be aware though that each camera format (size) has it's own 'normal' focal length which can be roughly deduced by measuring the diagonal of the format, e.g 50mm for a 35mm film camera, 300mm for an 8x10 inch camera

the focal length is an optical measure that doesn't change, whether it be normal, wide or telephoto depends on how much of the projected image is recorded by the format being used

Ray

A very good explanation. I recently created an 8x10 and utilized 1/2 of a rapid rectilinear lens whose FL was very close to 300mm. The scene was that of a country road which vanishes about a mile and a half in the distance. Viewing the contact print at a comfortable viewing distance (~ 12"-15") gives one the feeling of viewing the original scene through an 8"x10" window. The perspective is very normal.

 

[RobC]

interesting rob

but what about sensor size and angle of view

What about the sensor size? As you know some fish eye lenses project a cirlce onto the film area with a vignette, others are full frame. Depends on the lense design. The retina is not a flat plane like in a camera. And the question was, what will give the same near far relationships as the human eye. The answer is: The same focal length as the eye which is approx 16mm or 17mm.

 

[David A. Goldfarb]

The notion that the focal length of the lens should approximate that of the human eye used to be part of Minox's marketing rhetoric.

 

[Ole]

Focal length doesn't affect perspective.

Only lens placement affects perspective - and perceived perspective in a print is affected by the viewing distance: It looks natural when the picture fills the same angle of view as was used in the exposure. A wide-angle shot can look very natural if you make a BIG print and then look at it from a nose-lengths distance.

 

[Struan Gray]

Focal length doesn't affect perspective.

I know this to be true, and yet it is at the same time wholly false.

People don't look at pictures the way they look at scenes. Most people, certainly most photographers, can tell the difference between photographs taken with a normal lens, a lens with half the normal focal length and a lens with twice the normal focal length. They can make that distinction whatever the print size and however far away from them the print is held.

To me, that says we look at pictures and try to construct a world as seen by a normal lens that conforms with what we see. You can do as much ray-tracing and manipulation of mathematical formulea as you like, but *perceptually* different focal lengths do have different perspective.

So there.

 

[DannL]

Focal length doesn't affect perspective.

Only lens placement affects perspective - and perceived perspective in a print is affected by the viewing distance: It looks natural when the picture fills the same angle of view as was used in the exposure. A wide-angle shot can look very natural if you make a BIG print and then look at it from a nose-lengths distance.

A short focal length lens used to take a photograph at close range will cause wide-angle distortion.

 

[dpurdy]

I have always thought that vision is somewhat psychological or at least individual. There is a lot of information available to the eyes and a wide scope of vision but you choose how much of it to give attention to. (IMO) Some people see more telesopically and some more wide angle. I have always thought that partly accounted for different photography styles.

I use a fixed 80mm lens on my camera and have no frustration from the limitation because it seems to match how I look at the world. Chicken or the egg? I can't say but I have used the same type camera and lens for 30 years so maybe my vision is affected by the camera choice.

When I used to do a lot of 35mm photography I had a 24mm lens I would use occassionally and it always seemed to me that if I closed one eye and looked at everything my eye could see, that is about what I would get from my 24mm lens.

 

[RDKirk]

Normal is just "normal"

If you pick up a "normal" sized enlargement and view it from a "normal" distance, if the perpective looks "normal," you can bet money it was shot with a lens that was approximately the same focal length as the diagonal of the orginal format.

What's a "normal" sized enlargement viewed from a "normal"distance? It's probably expressible in terms of minutes of arc, but you can bet money it's about the same as an 8x10 print viewed from about 20 inches.

A lot of what we accept as "normal" is based on how much our eye/mind system sharply grasps in a single glance. That's why people will say that a "normal" lens approximates the view of the human eye. Technically, not. For instance, it's already been mentioned that optically the angle of view of the human eye is much wider than 45-50 degrees. But as has already been mentioned, the mind does a lot of "post processing" of the image to make it conform to past experience.

That's why if you fit more of a view that could be naturally grasped in a single glance into a photograph small enough to be grasped in a single glance, it will not look normal. The eye/mind "knows" it's never "seen" those perspective relationships before and rejects that view. "That's just wrong," the mind says of what they eye reports.

 

[dancqu]

For example, the image is upside down on the
retina but you don't see it that way.

And the brain enlarges the tiny image made
upon the retina. At least mine does. It makes
big things look big.

Someone mentioned ray tracing. I may work
with that. Dan

 

[Lee Shively]

I've avoided getting into this discussion since everyone seems to have his own definition of "perspective". But if anyone is interested, try to locate a copy of the old, out-of-print book Canon Lenses by George Lepp and Joe Dickerson and turn to pages 20-27. These pages have photographs taken with lenses from 15mm through 600mm. These pictures show how lenses affect perspective when photos are taken from the same position and also from different positions to maintain the same subject size.

 

[Lee Shively]

I clicked "post" too soon. I meant to add that the relationship between the main subject and the foreground/background is illustrated by the photos in the Canon book.

 

[Ray Heath]

What about the sensor size? As you know some fish eye lenses project a cirlce onto the film area with a vignette, others are full frame. Depends on the lense design. The retina is not a flat plane like in a camera. And the question was, what will give the same near far relationships as the human eye. The answer is: The same focal length as the eye which is approx 16mm or 17mm.

yes Rob, and if the sensor size was small enough to record only part of that projected circle then the resulting image may well not look particularly wide angle

of course sensor size must come into the equation

Ray